Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Letter
  • Published:

Correction of aberrant growth preserves tissue homeostasis


Cells in healthy tissues acquire mutations with surprising frequency. Many of these mutations are associated with abnormal cellular behaviours such as differentiation defects and hyperproliferation, yet fail to produce macroscopically detectable phenotypes1,2,3. It is currently unclear how the tissue remains phenotypically normal, despite the presence of these mutant cells. Here we use intravital imaging to track the fate of mouse skin epithelium burdened with varying numbers of activated Wnt/β-catenin stem cells. We show that all resulting growths that deform the skin tissue architecture regress, irrespective of their size. Wild-type cells are required for the active elimination of mutant cells from the tissue, while utilizing both endogenous and ectopic cellular behaviours to dismantle the aberrant structures. After regression, the remaining structures are either completely eliminated or converted into functional skin appendages in a niche-dependent manner. Furthermore, tissue aberrancies generated from oncogenic Hras, and even mutation-independent deformations to the tissue, can also be corrected, indicating that this tolerance phenomenon reflects a conserved principle in the skin. This study reveals an unanticipated plasticity of the adult skin epithelium when faced with mutational and non-mutational insult, and elucidates the dynamic cellular behaviours used for its return to a homeostatic state.

This is a preview of subscription content, access via your institution

Access options

Buy this article

Prices may be subject to local taxes which are calculated during checkout

Figure 1: β-Catenin mutant cells generate aberrant growths that regress.
Figure 2: Wild-type cells are required for the active elimination of mutant cells.
Figure 3: Fate of regressing tissue is niche-dependent.
Figure 4: Tissue correction occurs in mutant Hras model.

Similar content being viewed by others


  1. Ling, G. et al. Persistent p53 mutations in single cells from normal human skin. Am. J. Pathol. 159, 1247–1253 (2001)

    Article  CAS  PubMed  Google Scholar 

  2. Martincorena, I. et al. High burden and pervasive positive selection of somatic mutations in normal human skin. Science 348, 880–886 (2015)

    Article  ADS  CAS  PubMed  Google Scholar 

  3. Laurie, C. C. et al. Detectable clonal mosaicism from birth to old age and its relationship to cancer. Nat. Genet. 44, 642–650 (2012)

    Article  CAS  PubMed  Google Scholar 

  4. Pineda, C. M. et al. Intravital imaging of hair follicle regeneration in the mouse. Nat. Protocols 10, 1116–1130 (2015)

    Article  CAS  Google Scholar 

  5. Rompolas, P. et al. Live imaging of stem cell and progeny behaviour in physiological hair-follicle regeneration. Nature 487, 496–499 (2012)

    Article  ADS  CAS  PubMed  Google Scholar 

  6. Clevers, H. Wnt/β-catenin signaling in development and disease. Cell 127, 469–480 (2006)

    Article  CAS  Google Scholar 

  7. Liu, F. et al. β-Catenin initiates tooth neogenesis in adult rodent incisors. J. Dent. Res. 89, 909–914 (2010)

    Article  PubMed  Google Scholar 

  8. van Amerongen, R., Bowman, A. N. & Nusse, R. Developmental stage and time dictate the fate of Wnt/β-catenin-responsive stem cells in the mammary gland. Cell Stem Cell 11, 387–400 (2012)

    Article  CAS  Google Scholar 

  9. Harada, N. et al. Intestinal polyposis in mice with a dominant stable mutation of the β-catenin gene. EMBO J. 18, 5931–5942 (1999)

    Article  CAS  PubMed  Google Scholar 

  10. Gat, U., DasGupta, R., Degenstein, L. & Fuchs, E. De novo hair follicle morphogenesis and hair tumors in mice expressing a truncated β-catenin in skin. Cell 95, 605–614 (1998)

    Article  CAS  Google Scholar 

  11. Lo Celso, C., Prowse, D. M. & Watt, F. M. Transient activation of β-catenin signalling in adult mouse epidermis is sufficient to induce new hair follicles but continuous activation is required to maintain hair follicle tumours. Development 131, 1787–1799 (2004)

    Article  CAS  Google Scholar 

  12. Kretzschmar, K., Weber, C., Driskell, R. R., Calonje, E. & Watt, F. M. Compartmentalized epidermal activation of β-catenin differentially affects lineage reprogramming and underlies tumor heterogeneity. Cell Rep. 14, 269–281 (2016)

    Article  CAS  PubMed  Google Scholar 

  13. MacDonald, B. T., Tamai, K. & He, X. Wnt/β-catenin signaling: components, mechanisms, and diseases. Dev. Cell 17, 9–26 (2009)

    Article  CAS  PubMed  Google Scholar 

  14. Chan, E. F., Gat, U., McNiff, J. M. & Fuchs, E. A common human skin tumour is caused by activating mutations in β-catenin. Nat. Genet. 21, 410–413 (1999)

    Article  CAS  Google Scholar 

  15. Morin, P. J. et al. Activation of β-catenin-Tcf signaling in colon cancer by mutations in β-catenin or APC. Science 275, 1787–1790 (1997)

    Article  CAS  Google Scholar 

  16. Deschene, E. R. et al. β-Catenin activation regulates tissue growth non-cell autonomously in the hair stem cell niche. Science 343, 1353–1356 (2014)

    Article  ADS  CAS  PubMed  Google Scholar 

  17. Madisen, L. et al. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat. Neurosci. 13, 133–140 (2010)

    Article  CAS  Google Scholar 

  18. Ferrer-Vaquer, A. et al. A sensitive and bright single-cell resolution live imaging reporter of Wnt/β-catenin signaling in the mouse. BMC Dev. Biol. 10, 121 (2010)

    Article  CAS  PubMed  Google Scholar 

  19. Augustin, I. et al. The Wnt secretion protein Evi/Gpr177 promotes glioma tumourigenesis. EMBO Mol. Med. 4, 38–51 (2012)

    Article  CAS  PubMed  Google Scholar 

  20. Carpenter, A. C., Rao, S., Wells, J. M., Campbell, K. & Lang, R. A. Generation of mice with a conditional null allele for Wntless. Genesis 48, 554–558 (2010)

    Article  CAS  PubMed  Google Scholar 

  21. Pruitt, S. C., Freeland, A., Rusiniak, M. E., Kunnev, D. & Cady, G. K. Cdkn1b overexpression in adult mice alters the balance between genome and tissue ageing. Nat. Commun. 4, 2626 (2013)

    Article  ADS  PubMed  Google Scholar 

  22. Xie, W., Chow, L. T., Paterson, A. J., Chin, E. & Kudlow, J. E. Conditional expression of the ErbB2 oncogene elicits reversible hyperplasia in stratified epithelia and up-regulation of TGFα expression in transgenic mice. Oncogene 18, 3593–3607 (1999)

    Article  CAS  Google Scholar 

  23. Füllgrabe, A. et al. Dynamics of Lgr6+ progenitor cells in the hair follicle, sebaceous gland, and interfollicular epidermis. Stem Cell Rep. 5, 843–855 (2015)

    Article  Google Scholar 

  24. Snippert, H. J. et al. Lgr6 marks stem cells in the hair follicle that generate all cell lineages of the skin. Science 327, 1385–1389 (2010)

    Article  ADS  CAS  PubMed  Google Scholar 

  25. Baker, C. M., Verstuyf, A., Jensen, K. B. & Watt, F. M. Differential sensitivity of epidermal cell subpopulations to β-catenin-induced ectopic hair follicle formation. Dev. Biol. 343, 40–50 (2010)

    Article  CAS  PubMed  Google Scholar 

  26. Chen, X. et al. Endogenous expression of HrasG12V induces developmental defects and neoplasms with copy number imbalances of the oncogene. Proc. Natl Acad. Sci. USA 106, 7979–7984 (2009)

    Article  ADS  CAS  Google Scholar 

  27. Endo, M. et al. Efficient in vivo targeting of epidermal stem cells by early gestational intraamniotic injection of lentiviral vector driven by the keratin 5 promoter. Mol. Ther. 16, 131–137 (2008)

    Article  CAS  Google Scholar 

  28. Beronja, S., Livshits, G., Williams, S. & Fuchs, E. Rapid functional dissection of genetic networks via tissue-specific transduction and RNAi in mouse embryos. Nat. Med. 16, 821–827 (2010)

    Article  CAS  PubMed  Google Scholar 

  29. Beronja, S. & Fuchs, E. RNAi-mediated gene function analysis in skin. Methods Mol. Biol. 961, 351–361 (2013)

    Article  CAS  PubMed  Google Scholar 

  30. Means, A. L., Xu, Y., Zhao, A., Ray, K. C. & Gu, G. A CK19(CreERT) knockin mouse line allows for conditional DNA recombination in epithelial cells in multiple endodermal organs. Genesis 46, 318–323 (2008)

    Article  CAS  PubMed  Google Scholar 

  31. Tumbar, T. et al. Defining the epithelial stem cell niche in skin. Science 303, 359–363 (2004)

    Article  ADS  CAS  Google Scholar 

  32. Mesa, K. R. et al. Niche-induced cell death and epithelial phagocytosis regulate hair follicle stem cell pool. Nature 522, 94–97 (2015)

    Article  ADS  CAS  PubMed  Google Scholar 

  33. Fujiwara, H. et al. The basement membrane of hair follicle stem cells is a muscle cell niche. Cell 144, 577–589 (2011)

    Article  CAS  PubMed  Google Scholar 

  34. Fuchs, Y. et al. Sept4/ARTS regulates stem cell apoptosis and skin regeneration. Science 341, 286–289 (2013)

    Article  ADS  CAS  PubMed  Google Scholar 

  35. Kaufman, C. K. et al. GATA-3: an unexpected regulator of cell lineage determination in skin. Genes Dev. 17, 2108–2122 (2003)

    Article  CAS  PubMed  Google Scholar 

Download references


We thank K. Cockburn, K. Mesa, and S. Guo for feedback on the manuscript. We thank G. Panse, P. Myung, and C. Ko for assistance with brightfield images and clinical assessment of haematoxylin and eosin sections. We thank G. Gu for K19creER and M. Taketo for β-catnflox(Ex3)/+ mice. This work was supported by The New York Stem Cell Foundation and grants to V.G. by the Edward Mallinckrodt, Jr. Foundation, a Howard Hughes Medical Institute Scholar award, and National Institute of Arthritis and Musculoskeletal and Skin Disease, National Institutes of Health (NIH), grant numbers 5R01AR063663-04 and 1R01AR067755-01A1. The content is the responsibility of the authors and does not necessarily represent the official views of the NIH. S.Br. was supported by the NIH Predoctoral Program in Cellular and Molecular Biology, grant number NIH T32GM007223. C.M.P. was supported by Human Genetics Training Grant NIH T32HD001749 and is currently supported by the National Cancer Institute of the NIH under Award Number F31CA206419. S.P. was supported by a James Hudson Brown - Alexander Brown Coxe Postdoctoral Fellowship and is currently supported by CT Stem Cell Grant 14-SCA-YALE-05. V.G. is a New York Stem Cell Foundation Robertson Investigator. β-catnflox(Ex3) mice are available from M. Taketo under a material transfer agreement with Banyu-Merck. We thank the Marine Biological Laboratory, Woods Hole, Massachusetts, for support while writing this manuscript.

Author information

Authors and Affiliations



S.Br., C.M.P., and V.G. designed experiments, analysed data, and wrote the manuscript. S.Br. performed β-catenin imaging, immunohistochemistry, drug treatments, and mouse genetics. C.M.P. performed laser ablations, reverse transcribed PCR, Hras imaging, and viral work. T.X. assisted with β-catenin imaging and manuscript writing. J.B. assisted with immunohistochemistry and mouse genetics. K.S. assisted with clinical assessment and manuscript writing. S.P. assisted with drug delivery. C.M.-M. assisted with viral work. D.G. assisted with data analysis. J.R. provided training on viral work. S.Be. provided tools and feedback on the data and manuscript.

Corresponding authors

Correspondence to Slobodan Beronja or Valentina Greco.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Additional information

Reviewer Information Nature thanks J. Mills and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Figure 1 Increasing the number of mutant cells increases the severity of aberrancies.

a, Diagram of the increasing severity of aberrancies that develop as the tamoxifen dosage (and number of mutant cells) increases. bd, Representative hair follicles induced with increasing amounts of tamoxifen (TAM), upon induction (top), and their resulting phenotype (bottom): b, 10 μg of tamoxifen results in morphologically normal hair follicles with no outgrowths; c, 200 μg of tamoxifen induces about 3–15 recombined cells out of the roughly 80 HFSCs in the resting hair follicles, results in outgrowths; d, 6 mg of tamoxifen (1 mg of tamoxifen per day for 6 days) results in the formation of tumours. e, Quantification of outgrowths that arise from increasing number of labelled cells (mean ± s.e.m., n = 285 hair follicles in four mice). Scale bars, 50 μm.

Source data

Extended Data Figure 2 Developing mutant outgrowths are largely composed of a mixture of mutant and wild-type cells.

a, Percentage distribution of new bud composition, with 71% of developing outgrowths made up of a mixture of mutant and wild-type cells (n = 112 new buds across three mice). b, Representative budding outgrowths in tamoxifen-treated (200 μg) K19creER;β-catnflox(ex3) mice. All epithelial nuclei are labelled with K14–H2BmCherry (red). Nuclei of mutant cells are labelled with TCF–H2BGFP (green). All scale bars, 50 μm.

Source data

Extended Data Figure 3 Small population of mutant cells can generate a varying number of outgrowths that regress.

a, Wide-field view of emerging outgrowths. b, Outgrowth frequency per hair follicle generated by a small population of mutant cells (n = 718 hair follicles across three mice). c, Diagram of outgrowth regression.

Source data

Extended Data Figure 4 Mutant β-catenin cells differentiate and are subsequently lost from the tissue.

a, Whole-mount β-catenin immunofluorescence staining (blue) showing increased nuclear β-catenin staining corresponds to the Cre and Wnt fluorescent reporters. b, β-Catenin whole-mounts. tdTom+ cells (red) along the core of the outgrowth express higher levels of β-catenin (green). c, Revisits of the same outgrowth, showing the progressive elimination of the mutant core (TCF–H2BGFP, green). Epithelial nuclei are labelled with K14–H2BmCherry (red). d, Reverse transcribed PCR quantification of mRNA expression of the floxed mutant β-cateninΔex3 allele, normalized to GAPDH, during the development and regression of outgrowths (mean ± s.e.m., n = 3 mice per time point, ****P < 0.0001). e, Whole-mount Ki67 staining. f, Mitotic events in tdTomato negative (tdTom) and positive (tdTom+) cells, during different outgrowth stages (mean ± s.e.m., n = 150 outgrowths across three mice). g, Revisit of a single outgrowth showing the morphologically differentiated core (arrowheads, left) that is expelled from the tissue over time (right). h, Whole-mount staining of inner root sheath differentiation marker Gata3 (ref. 35) co-localizes with mutant core labelled with tdTomato (left) and TCFH2BGFP (right). Scale bars, 50 μm.

Source data

Extended Data Figure 5 Apoptosis and cell extrusion occurs during both growth and regression of aberrancies.

a, Quantifications of different cellular behaviours occurring during growth and regression (mean ± s.e.m.). b, Representative outgrowths during growth (top, lower left) and regression (lower right) showing nuclear fragmentation, a morphological hallmark of apoptosis, pseudocoloured pink. c, Examples of outgrowths during growth (top) and regression (bottom) showing epithelial cells extruded into the dermis (pseudocoloured green). d, Revisits of extruded cells in the dermis, showing cells become undetectable over time. e, Quantifications showing the amount of time cells are detectable in the dermis once extruded. All scale bars, 50 μm.

Source data

Extended Data Figure 6 Outgrowths with limited contribution of wild-type cells develop into proliferative cysts.

a, K19creER;B-catnflox(Ex3);tdTomato Wlsfl/fl whole-mount of Ki67 staining. b, K19creER;β-catnflox(Ex3);K14rtTA;cdkn1b whole-mount of Ki67 staining. c, Haematoxylin and eosin image of K19creER;β-catnflox(Ex3) backskin at low- (left) and high-power (right) magnifications. d, Haematoxylin and eosin image of backskin from K19creER;β-catnflox(Ex3);Wlsfl/fl showing the formation of small follicular cysts at low- (left) and high-power (right) magnification. e, Haematoxylin and eosin image of backskin from K19creER;β-catnflox(Ex3);K14rtTA;cdkn1b showing a dilated follicular cyst at low- (left) and high-power (right) magnifications. All scale bars, 50 μm.

Extended Data Figure 7 Endogenous and ectopic cellular behaviours are used in the regression of the outgrowth.

Diagram of observed cellular behaviours during the growth and regression of the outgrowths.

Extended Data Figure 8 Remaining cells from regressed outgrowths can develop into functional appendages.

a, Revisits of the same outgrowth in the non-cycling compartment. K14–H2BGFP labels epithelial nuclei (green). The non-hair cycling compartment is outlined in white. The cycling hair follicle, out of focal plane at later time points, is represented by a white dotted line. Insets show elimination of mutant core (tdTomato, red). b, Quantification of the phases of growth and regression of the outgrowths from the hair cycling epithelium of tamoxifen-treated (200 μg) Lgr6creER;β-catnflox(Ex3)/+;tdTomato;mTmG;K14–H2BGFP mice (mean ± s.e.m.). Pre-phenotypic plot accounts for hair follicles before the development of outgrowths to demonstrate variability in timing of formation (n = 33 outgrowths in three mice). c, Quantification of the two different fates of non-cycling epithelial outgrowths 173 days after induction (n = 24 outgrowths). d, Revisits from a tamoxifen-treated (200 μg) Lgr6creER;β-catnflox(Ex3)/+;tdTomato;mTmG;K14–H2BGFP mouse showing that, after the mutant core has differentiated out of the tissue (left), the remaining nodule of cells has converted into an extranumerary hair follicle. Subsequent revisits capture the representative extranumerary follicle in the early stage of catagen (centre) and telogen (right). e, Diagram representing the conversion into extranumerary structures. f, Revisits of representative Lgr6creER;β-catnflox(Ex3)/+;tdTomato;mTmG;K14–H2BGFP mouse after the administration of a high-dose (2 mg) of tamoxifen (n = 4 mice). Large benign tumours result in a macroscopic phenotype of severely wrinkled skin (day 21, left and inset) that resolves over time as the outgrowths regress (day 208, right and inset). g, Quantification of the hair follicles with tumours (mean ± s.e.m., n = 803 hair follicles across three mice). h, Quantification of hair follicles with supernumerary sebaceous gland lobes and de novo hair follicles after outgrowth regression (mean ± s.e.m., n = 580 hair follicles across three mice). i, Representative images of normal (left) and extranumerary (right) sebaceous glands present in the tissue 151 days after induction. Normal sebaceous glands are outlined in white, extranumerary structures are outlined in blue. j, Oil Red O staining on tail whole-mounts of Lgr6creER;β-catnflox(Ex3)/+ mouse 1 year after the administration of a high dose (2 mg) of tamoxifen. Note the Oil Red O staining in the extranumerary sebaceous gland lobes, indicating these structures are functional and actively producing lipids. k, Whole-mount Pcad IF staining on Lgr6creER;β-catnflox(Ex3)/+ ear skin 1 year after the administration of a high dose (2 mg) of tamoxifen, showing the extranumerary hair follicle has a Pcad enrichment at the base of the structure, similar to the hair germ of the main axis. All scale bars, 50 μm.

Source data

Extended Data Figure 9 Broad Hras activation and mutation-independent systems reveal a wide range of corrective abilities.

a, Representative phenotypes in LV-cre HrasG12V broad activation model (59% are hyperthickened and 10% are structurally deformed). b, Quantification of follicular bulge width demonstrates that Hras follicles are significantly hyperthickened during the first rest (P < 0.0001) and growth phases (P < 0.0001), but that hyperthickening is resolved by the second rest phase (P = 0.0860, n = 492 wild-type follicles, 550 Hras follicles, three mice per genotype). c, Haematoxylin and eosin images demonstrating regions of epidermal hyperplasia (upper arrowhead) and dermal cysts (lower arrowhead). d, Keratinized cyst resolves over time. e, Top: xy view of macroscopic growth in Hras mouse demonstrates that hyperthickening and hyperkeratinization are resolved in 6 weeks. Hair follicles (A–C) are used as landmarks. Bottom: yz view of macroscopic growth at starting (p38) and ending (p80) time points (n = 3 regressing macroscopic growths observed in two mice).

Source data

Extended Data Figure 10 Non-mutational insult to wild-type follicles can be corrected.

a, Correction of a deformed hair follicle post-ablation (right) compared with its non-ablated counterpart (left). b, Frequency of ablated follicle recovery (mean ± s.e.m., n = 38 follicles in five mice) c, Representative image of an ablated follicle that adopts a deformed structure, corrects it, and begins to cycle again. d, Quantification of the cycling ability of follicles 4 weeks post-ablation revealed 24% of ablated follicles were able to cycle again (mean ± s.e.m., n = 38 follicles in five mice). Asterisks denote region of autofluorescence produced post-ablation. All scale bars, 50 μm.

Source data

Supplementary information

Video 1: The mutant core differentiates out of the tissue through the epidermis

A z-stack beginning at the level of the outgrowth, first moving up to the epidermis, then back down to the outgrowth of a Tamoxifen-treated (200µg) K19CreER;β-catnflox(Ex3)/+;tdTomato/+;K14-H2BGFP;Lef1RFP mouse. Epithelial cells are labelled with K14-H2BGFP (green), black regions are indicative of mosaicism of the transgene. Fibrillary collagen I visualized via secondary harmonic signal in blue. Note the mutant tdTomato-labeled core of the outgrowth (red) spans the length of the outgrowth and snakes out of the hair follicle through the epidermis and emerging from the skin. Step size = 3 µm. All scale bars in figure = 50 µm.

Video 2: Homeostatic and ectopic cellular behaviors occur in a regressing outgrowth

Timelapse of a regressing outgrowth of a Tamoxifen-treated (200µg) K19CreER;β-catnflox(Ex3)/+;tdTomato;K14-H2BGFP mouse. Insets visualized with K14-H2BGFP. Upper inset shows nuclear fragmentation during apoptosis. Lower inset shows ectopic behavior of retrograde migration of suprabasal cells, back into main axis of hair follicle. Note lower inset displayed at higher frame rate in order to appreciate movement of cells (h:mm:ss.fff).

Video 3: Extrusion of a cell from the outgrowth into the dermis during regression

A six hour timelapse of a regressing outgrowth in a Tamoxifen-Brown* and Pineda* et al, 2017; Correction of aberrant growth treated (200µg) K19CreER;β-catnflox(Ex3)/+;K14-H2BGFP mouse. Visualized using K14-H2BGFP (h:mm:ss.fff). Note the differentiated mutant core has already been eliminated from the tissue. Arrowhead highlights the cell being extruded from the tip of the outgrowth into the surrounding dermis. Scale bar = 20 µm.

Video 4: Proliferation and cell death occur simultaneously in the growing outgrowth

A 4-hour timelapse of a growing outgrowth in Tamoxifen-treated (200µg) K19CreER;β-catnflox(Ex3)/+tdTomato; ;K14-H2BGFP mouse, showing apoptosis is occurring concurrently with proliferation. (h:mm:ss.fff)

Video 5: Wild-type cells fail to enter outgrowths when proliferation is blocked

Z-stacks, beginning at the epidermis and moving down to the outgrowth in Tamoxifen-treated (200µg), K19CreER;β-catnflox(Ex3)/+tdTomato; K14-H2BGFP;Lef1RFP mouse (left) compared to a Tamoxifen-treated (200µg)/Doxycycline-treated (2mg/ml) K19CreER;β-catnflox(Ex3)/+tdTomato;K14rtTA; TetO-cdkn1b ;K14-H2BGFP mouse (right). Note all the layers in the β-catenin/CDKN1b outgrowth are composed of tdTomato-labeled mutant cells, while only a few wild-type cells can be seen at the base. Step size = 3 µm. Scale bar = 20 µm.

PowerPoint slides

Source data

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Brown, S., Pineda, C., Xin, T. et al. Correction of aberrant growth preserves tissue homeostasis. Nature 548, 334–337 (2017).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing